Abstract: Tuberculosis is an epidemic disease that points to death. It spreads through the air by person anguish from Tuberculosis. Ten or more people can be infected per year by a single patient. Like any transmittable disease, TB is ubiquitous even in urbanized nations. But it is a foremost problem in the established nations including India. Soft computing is an agglomeration of methodologies. For tuberculosis detection, consolidation of the Genetic algorithm and Neural Network is proposed. This soft computing technique is used to acquire an optimized solution that is needed to seize correct decision for improved performance and gave the rigorous result. The main aspiration of this scrutiny is to pre-predict information about tuberculosis disease at low outlay. Genetic Algorithm has Fitness function, Crossover and Mutation operations to disentangle optimization problems. In this study, various parameters are taken as an input and conform five hidden nodes afterward hatch an output. The Neural Network is trained with Genetic Algorithm and Back propagation algorithm to improve accuracy. There are ten parameters such as Age, Gender, Cough, Fever, Chest Pain, Weight Loss, Night Sweats, and Unwillingness for work and Loss of Appetite. These parameters are adopted by many researchers to identify Tuberculosis.
Keywords: Tuberculosis, Genetic Algorithm, Neural Network, Soft Computing, Data Mining.